摘要
本文对147例胃癌根治术患者,运用Cox模型从26个预后因子中,确定浸润型生长方式、淋结巴转移、幽门部肿瘤、手术范围、年龄、症状为胃癌预后因子.其中,以浸润型生长方式与术后生存期的关系最为密切,淋巴结转移是仅次于生长方式的重要因素。根据预测模型对每个患者都可计算预测值(PV),判断预后。其敏感度为90.4%,可靠性达87.7%,高于其他方法的预测效果。生存函数估计则可提供更为详细的术后生存信息,如1年、2年、5年、10年生存率和中位生存期,进而可估测术后生存期,为临床医师选择治疗方案和术后监护提供依据。上述各项分析由我们在长城0520微型计算机上编制的生存分析软件系统实现。
This paper presented the first time application ofCox regression model in survival analysis of gastriccancer in China.Using Cox model,a multivariateanalysis was made from 147 patients with gastriccarcinoma treat by courative resection.Among thetwenty-six factors,only infiltrating tumor growthpattern,lymph node metastases,tumor located in an-trun,age,type of operation and the number of sym-ptoms showed a high significance with prognosticrelationship.Infilttrating growth pattern is most im-portant■factor which influenced postoperative survivaltime.Lymph node metastases presented second impo-rtant role.The predicting value(PV)may calculatedaccording to patient's individual fea■uers,which canbe used as evaluation of prognosis.It's sensitivityand reliability achieved 90.4% and 87.7%,respect-ively.Furthermore,the estamation of n year survivalrate may be from the survival function.Those pre-diction could provide useful information for surgeonto treat and follow up patients.
出处
《中国卫生统计》
CSCD
北大核心
1989年第1期34-37,共4页
Chinese Journal of Health Statistics
关键词
胃癌
预后
COX模型
多因素分析
Prognosis of gastic cancer
Cox regression model
Multivariate analysis
